We examine the antecedents and outcomes of new ventures' formation of multilateral R&D alliances. Our results show an inverted U-shaped relationship between market uncertainty and a new venture's likelihood of forming multilateral R&D alliances. Top management team's social capital and ventures' technological capabilities are critical for new ventures to identify and capture alliance opportunities. Moreover, our analysis reveals value creation effects of multilateral R&D alliances for new ventures despite the challenges and difficulties associated. We further show that the value creation effect is a function of the type of exchange relationship (i.e., net- vs. chain-based) in the multilateral R&D alliance and that governance structure moderates this relationship.
While scholars have studied the benefits and risks of bilateral alliances for new ventures, the antecedents and consequences of ventures' engagement in multilateral R&D alliances are under-researched. As a widely dispersed organizational form,1 multilateral alliances offer partner firms access to complementary resources, market information, and investment opportunities (e.g., Mitchell et al., 2002, Sakakibara, 2002 and Yin and Wu, 2003). These benefits can be critical to venture survival and growth. However, significant amount of risk stems from the complicated alliance design and governance associated with multilateral alliances (Doz and Hamel, 1998). If not managed and protected carefully, valuable knowledge owned by new ventures can be appropriated. There has been a great deal of research on new ventures' involvement in alliances in organization theory (e.g., Alvarez and Barney, 2001, Arino et al., 2008, Deeds and Hill, 1996, Gomes-Casseres, 1994, Larson, 1992 and Stuart et al., 1999) and increasing amount of research on multilateral alliances from the incumbents' perspectives (e.g., Das and Teng, 2002, Lavie et al., 2007 and Li et al., forthcoming). Yet, our understanding of why new ventures opt to take on the challenges and dive into the complicated cooperative relationship involving multiple partners and whether market value is created by multilateral R&D alliances for the venture partners remains restricted.
Descriptive statistics for the first- and second-stage models are reported in Table 2 and Table 3. We carefully examine our data for potential multicollinearity problems. None of the absolute values of correlations is above 0.50. We check the variance inflation factors (VIF) for all individual variables which are below 5 and the average VIFs for all regression models below 2. Both indices are below the accepted thresholds of 10 and 2, respectively (Neter et al., 1990). Therefore, we are confident that our estimates are not threatened by multicollinearity problems.
